Power sharing, peace and democracy: Any obvious relationships?

Internationalhttp://ias.sagepub.com/ Area Studies Review Power sharing, peace and democracy: Any obvious relationships?...

Author: Binningsbø | Helga Malmin

18 downloads 571 Views 494KB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form

DOWNLOAD PDF

Internationalhttp://ias.sagepub.com/ Area Studies Review

Power sharing, peace and democracy: Any obvious relationships? Helga Malmin Binningsbø International Area Studies Review 2013 16: 89 DOI: 10.1177/2233865912473847 The online version of this article can be found at: http://ias.sagepub.com/content/16/1/89

Published by: Center for International Area Studies, Hankuk University of Foreign Studies

Peace Research Institute Oslo

and http://www.sagepublications.com

Additional services and information for International Area Studies Review can be found at: Email Alerts: http://ias.sagepub.com/cgi/alerts Subscriptions: http://ias.sagepub.com/subscriptions Reprints: http://www.sagepub.com/journalsReprints.nav Permissions: http://www.sagepub.com/journalsPermissions.nav Citations: http://ias.sagepub.com/content/16/1/89.refs.html

>> Version of Record - Mar 7, 2013 What is This?

Downloaded from ias.sagepub.com at Universitet I Oslo on October 28, 2014

473847 2013

IAS16110.1177/2233865912473847International Area Studies ReviewBinningsbø

Review Essay

Power sharing, peace and democracy: Any obvious relationships?

International Area Studies Review 16(1) 89–112 © The Author(s) 2013 Reprints and permissions: sagepub.co.uk/journalsPermissions.nav DOI: 10.1177/2233865912473847 ias.sagepub.com

Helga Malmin Binningsbø Peace Research Institute Oslo, Norway

Abstract Power sharing, mostly understood as including political opponents in a joint executive coalition government, is today a dominant approach to solving conflict. Almost as a panacea it has been introduced in numerous war-affected countries and is often recommended as a political solution to overcome deep divisions between groups. Researchers, mediators and policy-makers applaud such solutions as forward-looking, peace-strengthening and democratic. However, many have criticized power sharing and its failed ability to create peace and development in divided and conflict-ridden countries. The literature on power sharing can reach such divergent conclusions because there is no consensus on what power sharing is, what the aim of it is and how to study it. Apart from broad inclusion in joint government, the understanding of power sharing varies and recommendations to ‘share power’ give little guidance to policy-makers aiming to mitigate conflict. Therefore, this article reviews 40 years of research on power sharing by elaborating on four central aspects within the literature: (1) conceptualization; (2) domain; (3) causal mechanism; and (4) measurement. While there may be no ‘true’ power sharing or ‘truth’ about power sharing, the article concludes that it is crucial that researchers and policy-makers are clear about which type of power sharing they are discussing in specific situations. Given that power sharing is increasingly recommended and implemented in many fragile and post-conflict societies, it is important to understand what is meant by power sharing in these contexts. Keywords Civil war, consociationalism, democracy, peace, power sharing

Introduction Power sharing, mostly understood as including political opponents in a joint executive coalition government, is today a dominant conflict-solving approach. Almost as a panacea it has been introduced in numerous conflict-ridden countries. Power sharing is a widely used tool and has been Corresponding author: Helga Malmin Binningsbø, Peace Research Institute Oslo, Hausmannsgate 7, Oslo, 0186, Norway. Email: [email protected]


90

International Area Studies Review 16(1)

applied not only to ease ethnic tensions following the December 2007 election in Kenya or to mitigate political conflict between Tswangerai’s Movement for Democratic Change and Mugabe’s Zimbabwe African National Union (Patriotic Front) in Zimbabwe, but also to terminate civil wars in Bosnia, Burundi, Cambodia, Lebanon, Nepal and Sierra Leone, and perhaps the most successful and known case – the termination of the apartheid regime in South Africa (Sisk, 1995). However, apart from broad inclusion in joint government, the understanding of power sharing varies and simply suggesting sharing power gives little guidance to policy-makers aiming at mitigating conflict.1 The aim of this article is to make the understanding of ‘power sharing’ less confusing. By discussing the different conceptualizations, causal chains and analytical approaches presented during 40 years of power sharing research, it gives researchers and policy-makers a better knowledge base for their own choices. Although power sharing is a widely used and often recommended political solution to overcome deep divisions between groups, many have criticized it and its failed ability to create peace and development in divided and conflict-ridden countries (e.g. Horowitz, 1985; Mehler, 2009; Roeder and Rothchild, 2005; Spears, 2000, 2002; Sriram and Zahar, 2009; Tull and Mehler, 2005). Other researchers applaud such solutions as forward-looking, peace-strengthening and democratic (Hartzell and Hoddie, 2007; Lijphart, 1977; Mukherjee, 2006a, b; Norris, 2008; Walter, 2002). The literature on power sharing can reach such divergent conclusions because there is no consensus on what power sharing is, what the aim of it is and how to study it. As a governing system, power sharing is mainly studied, suggested and implemented with two outcomes in mind: democracy and peace. However, both perspectives assume that the societies in need of such governance are conflictual – violent or not – thus, in both cases the aim of power sharing is to avoid conflict and ultimately to achieve peace. These relationships between power sharing, democracy and peace are central in the literature. However, power sharing research differs substantially both on definitions of the three concepts (power sharing, democracy and peace) and which causal chain it theorizes about (power sharing to democracy, power sharing to peace, democracy to peace and peace to democracy). Additionally, the research differs substantially on how it analyses these concepts and their relations to potential outcomes. Table 1 summarizes recent research on power sharing. The table shows the lack of consensus on what it is, its aim and the best way to study it. Some researchers view power sharing as specific provisions within peace agreements; some view it as specific governing institutions; some see power sharing as improving democracy; some see it as hindering democracy; some see it as stabilizing peace; some see it as reproducing violence, while some reach inconclusive findings about its consequences. Some researchers analyse power sharing statistically, while others study only one case. Some look at all civil wars, while others look at only peace agreements. Some researchers study the entire time period since the Second World War, while others look only at the years after the end of the Cold War. There is clearly no agreement about the answer to the questions ‘what is power sharing?’ and ‘how does it work?’ Rather, the answer depends on the writer’s (and reader’s) preferences. This article reviews different answers to these questions and gives an overview of some of the main understandings of power sharing, explaining how these differences influence the research and conclusions drawn. In particular, the article elaborates on four aspects of power sharing research: (1) conceptualization; (2) domain; (3) causal mechanism; and (4) measurement.

Concept: Broad arrangements or specific provisions? Power sharing is a contested concept. There are many meanings and definitions associated with it. A large variety of conceptualizations cloud our understanding of power sharing. On the one hand, Downloaded from ias.sagepub.com at Universitet I Oslo on October 28, 2014

Domain

49 negotiated settlements (38 peace agreements), 1945– 1999

Cyprus (pre-1960– 1974), New Zealand (1867–1998)

83 peace agreements, 1989–2004

South Africa, postapartheid

Research

Hartzell and Hoddie (2007)

Jarstad (2001)

Jarstad and Nilsson (2008)

Jung and Shapiro (1995)

Power sharing as inclusive institutional structures that do not provide for official opposition

Power sharing as peace agreement provisions for inclusion of rebels

Consociationalism understood as two underlying theoretical dimensions: inclusion and decentralization


Concept

Table 1. A selection of research on power sharing.

Implementing military and territorial power sharing entails costly concessions and should prevent conflict recurrence, but implementation of political power sharing has low costs and thus no effect on postconflict peace Power sharing prevents a vibrant opposition. Without opposition, citizens lack the ability to change government through elections, thus power sharing is undemocratic

Power sharing along many dimensions provide influence, are reinforcing and signal willingness to recognize and cooperate with a former enemy Inclusion levels the power balance between groups and decentralization removes conflict issues from national politics. These mechanisms reduce the level of conflict

Causality

PR electoral law and proportional distribution of cabinet portfolios in 1993 interim constitution: descriptive

Provisions and/or implementation of political, military and territorial power sharing: dummy variables

Settlement institutionalization (political, military, territorial, economic power sharing): count variable Ethnic quota systems in Parliament: descriptive

Measurement


(Continued)

Power sharing guarantees do not provide a long-term basis for democracy

A fixed ethnic quota system did not prevent state collapse in Cyprus, while a flexible ethnic quota system reduced tensions in New Zealand Territorial pacts and the implementation of military and territorial power sharing pacts decrease risk of peace failure

Power sharing prolongs post-conflict peace

Finding

Binningsbø 91

Austria (1945–1966), Belgium (1918–1963), the Netherlands (1917– 1967), Switzerland (1959–1977)

46 peace agreements, 1945–2005

19 African peace agreements, 1999–2007

Mattes and Savun (2009)

Mehler (2009)

Rwanda, Burundi, Democratic Republic of Congo (DRC), postCold War

Lemarchand (2007)

Lijphart (1977)

Domain

Research

Table 1. (Continued)




Consociational arrangements: grand coalition, mutual veto, proportionality, segmental autonomy

Power sharing as ad hoc concessions giving opposition a stake in democratic transitions, consociationalism as carefully designed constitutional norms

Concept

Consociational institutions help segment leaders to cooperate and reach moderate policies. Over time, this elite cooperation will trickle down to cooperation at the mass level and create stable democracy Power sharing provisions put constraints on an opponent’s ability to back out from an agreement and thus lessen insecurity and fear, reducing the likelihood of conflict recurrence Power sharing at national elite level does not address security concerns of civilians

Power sharing works better in transitions to democracy if it contains consociational traits; however, the (political) context in which it is implemented is crucial

Causality

Provisions of political, military, territorial and economic power sharing: descriptive

Provisions of political, military, territorial and economic power sharing: dummy variables

Power sharing failed to establish democracy in Rwanda, partly failed in the DRC, but perhaps more successful in Burundi

Power sharing and consociational provisions in Rwanda (1993 agreement), Burundi (2005 constitution), DRC (2002 agreement): descriptive Grand coalition, mutual veto, proportionality, segmental autonomy: descriptive

Elite focus of power sharing deals hinders civil society inclusion and democracy

Political power sharing prolongs peace duration

Consociationalism creates stable democracy

Finding

Measurement

92 International Area Studies Review 16(1)

Domain

111 civil wars, 1946–1999

111 civil wars, 1944–1999

191 countries, 1970–2004

36 stable democracies, 1970–1999

658 ethno-political dyads, 1955–1999

Research

Mukherjee (2006a)

Mukherjee (2006b)

Norris (2008)

Reilly (2005)

Roeder (2005)



Inclusion of ethnopolitical leaders in government

Government arrangement that gives significant ethnic groups access to decision-making bodies

Broad-based and inclusive political institutions

Peace agreement provisions for inclusion of rebels through political power sharing

Broad-based and inclusive political institutions

Concept

Power sharing ensures minority representation in decision-making processes and creates incentives for cooperation Inclusion encourages moderation and joint problem solving, which is important for successful democratization as well as sustaining democracy in the long run Power sharing institutions strengthen ethnic divisions and create incentives to escalate conflict

Inclusive political institutions make belligerents confident they will not be excluded from decision-making, thus reducing risk of conflict recurrence Power sharing together with military victory convinces combatants they are unable to continue violence

Causality

Parliamentarism, territorial autonomy, secondorder autonomy, disfranchisement, non-ethnic federalism: dummy variables

Oversized coalitions, PR elections: dummy variables

Parliamentarism, PR elections, federalism, free press: dummy variables

Political power sharing: dummy variable

Parliamentarism, PR elections: dummy variables

Measurement

(Continued)

Dyads with power sharing are more likely to see ethnopolitical conflict escalate to crisis and armed struggle

Ethnically divided democracies have executive inclusion, but not necessarily PR elections

Political power sharing and military victory prolong peace; political power sharing and stalemate shorten peace Power sharing strengthens democracy

Parliamentarism and PR elections prolong post-conflict peace

Finding

Binningsbø 93

Domain

Sri Lanka, Sudan, and Colombia, post-Cold War

Democratic Republic of Congo, post-Cold War

72 civil wars, 1944– 1992

Research

Sriram (2008)

Tull and Mehler (2005)

Walter (2002)



Peace agreement provisions for inclusion of rebels through political power sharing

Various governing initiatives which include armed groups in structures of power

Concept

Guaranteed representation of belligerents in government prevents former enemies from capturing the state and reduces risk of conflict recurrence

Power sharing creates incentives for groups craving power to use violent means

Power sharing builds on and strengthens cleavages between groups, increasing the risk of conflict

Causality

Provisions of political power sharing in 1999 and 2002 peace agreements: descriptive Provisions of political, military and territorial power sharing: dummy variables

Political, military, territorial and economic power sharing: descriptive

Measurement

Political and territorial power sharing increase likelihood of successful settlement

Power sharing was not offered in Sri Lanka and Colombia, and failed to be implemented in Sudan Political power sharing rewards violence and reproduces insurgency

Finding

94 International Area Studies Review 16(1)


95

Binningsbø

there are those that study one particular type of arrangement (Mukherjee, 2006b; Reilly, 2005; Reynolds, 1999), while on the other side of the spectrum are researchers employing a wide understanding of power sharing. Norris (2008), for example, includes a free press in her understanding of power sharing, while Gates and Strøm (2013) include electoral commissions where political parties are barred from office as a type of dispersed power sharing. Additionally, some understand power sharing as a specific political pact between former warring parties about legislative and/or executive power (Jarstad, 2008; Walter, 2002), while others include military, territorial and economic power sharing as well (Hartzell and Hoddie, 2007; Mattes and Savun, 2009; Mehler, 2009; Sriram, 2008). Below I elaborate on these different conceptualizations. Within the power sharing literature the works of Arend Lijphart (1969, 1977, 1985, 1999) are central. The Dutch-born political scientist was one of the first to describe and theorize about consociational democracy in divided societies, and he is seen as the father of power sharing – even baptized ‘Mr Consociation’ by the late Stein Rokkan (Lijphart, 2002: 37). Lijphart’s interest in power sharing emerged from studying the political system of plural European societies. He first used the term ‘consociational democracy’ in the late 1960s (1968, 1969).2 His concept had two dimensions: a socio-political one, referring to the pluralistic character of the society, and a political dimension, referring to the political cooperation of the elites (Lijphart, 1977: 5; Bogaards, 2000: 399). Elite cooperation is characterized by four institutional features: grand coalition, mutual veto, proportional representation system and segmental autonomy (Lijphart, 1977: 25). According to Lijphart, the most important of these consociational features is the grand coalition. In it, leaders of all significant segments shall rule together, searching compromise and consensus. These leaders are supposed to represent ‘all major linguistic and religious groups’ (Lijphart, 1996: 258) and/or ‘all of the important rival groups’ (Lijphart, 1998: 101). The grand coalition may take different forms: such as that of a grand coalition cabinet in parliamentary systems, a grand coalition of a president and other top officeholders in presidential systems, and broadly inclusive councils or committees with important advisory and coordinating functions. (Lijphart, 1985: 7)

The second aspect of Lijphart’s conceptualization of power sharing, mutual veto, is an instrument to assure political protection of minorities through the ability to block political decisions (Lijphart, 1977: 36; Schneckener, 2002: 205). A mutual veto does not have to be ultimate; rather it can be used as a means to obtain more time to deal with the questions involved, as a delaying veto. It can also be indirect, such that a certain quota of the coalition members has to agree to a decision for it to be approved. Consociational democracy is further based on a proportional distribution of influence (Lijphart, 1977: 40). The most common use of the proportionality principle is as an instrument to distribute seats in the legislature. However, it is also a means of allocating civil service and judicial appointments and financial resources, and assuring an adequate representation within the army and state-owned companies (Lijphart, 1977: 38ff.; Schneckener, 2002: 205). Finally, Lijphart (1977) argues that it can be wise to leave as many decisions as possible concerning the different segments to each respective population. Issues of segments’ exclusive concern include, among others, questions about religion, language and education. Such a segmental autonomy can be based on either group membership or territory, depending on the demographic distribution of the people. Where the regional cleavages correspond to the segmental cleavages, this may take the form of federalism, while group rights are more common when settlement patterns are dispersed. In sum, central to Lijphart’s concept of power sharing


96


is government by elite cooperation and consensus, in the form of grand coalition, mutual veto, proportionality and segmental autonomy. Norris (2008) also uses the words ‘power sharing’ to describe the type of governance she studies. However, her conceptualization is quite different from Lijphart’s consociational power sharing (Lijphart, 1969, 1977). Rather, her theoretical approach resembles Lijphart’s (1999) broader theory about consensus democracy, in particular, through Norris’s analytical contrasting of power sharing and power concentrating regimes. Norris (2008) studies four institutions: executive type (parliamentarism vs presidentialism), electoral systems (proportional representation vs majoritarian systems), vertical centralization (federalism vs unitary states) and independent mass media. Both Lijphart (1969, 1977, 1999) and Norris (2008) apply broad concepts of power sharing, including different decision-making bodies as characteristics of societies that share power. Other researchers look more exclusively at a few specific institutions. In particular, the proportional electoral system has received substantial attention. In some instances, proportional representation (PR) elections are considered beneficial because broad representation in parliament or other decision-making bodies is an aim in and of itself, especially in plural societies with diverging interests (Reynolds, 1999). In other instances, PR electoral laws are valuable because they encourage broad-based coalition governments, equally important for divided societies (Reynal-Querol, 2002; see, however, Reilly, 2005). Jarstad’s (2001) conceptualization of consociationalism is as a theory along two dimensions: inclusion and decentralization. She investigates one particular institution that affects the distribution of seats in legislature, namely ethnic quota systems. Within the growing body of literature on post-conflict power sharing, most studies look at provisions in peace agreements (Hartzell and Hoddie, 2007; Jarstad, 2008; Jarstad and Nilsson, 2008; Lemarchand, 2007; Mattes and Savun, 2009; Mehler, 2009; Mukherjee, 2006a; Sriram, 2008; Walter, 2002). Most of these studies have the same broad approach to power sharing as Lijphart (1977) and Norris (2008), distinguishing between different types of power sharing: political, military, territorial and economic. Political power sharing normally means provisions that guarantee rebel representatives positions of power (Jarstad, 2008; Walter, 2002).3 This representation can occur in a cabinet – a ‘government of national unity’ – but also in the legislature or other governing bodies. As such, political power sharing as understood in post-conflict studies resembles the consociational grand coalition (Lijphart, 1977). Distribution of seats in cabinet is often restricted in time, such as the ‘sunset clause’ in the Lomé agreement for Sierra Leone, which guaranteed a power sharing cabinet only until the next election, although some peace agreements include quotas for representation in parliament, often involving constitutional reform (Gilbert, 2009). Lijphart (1977) briefly discusses proportionality in the army as something that can be suitable for divided societies, but military power sharing can be crucial in post-conflict countries. When rebels are included in a (new) national army, it is viewed by both actors as a costly and credible signal about commitment to the peace agreement, and thus it is less likely that violence will escalate (Hoddie and Hartzell, 2003). Additionally, including former rebels in the army provides them with a job and perhaps a minimal income. Young men and women who have fought as insurgents, perhaps for many years, often lack education and/or skills training, and military integration may therefore serve as an alternative economic opportunity for former combatants (Glassmyer and Sambanis, 2008). Many armed conflicts are territorial, with rebels from a certain region fighting to achieve more autonomy, maybe even independence. To meet such demands, peace agreements can contain territorial power sharing. With this, some regions are guaranteed more power to decide over issues of their own concern. Federalism is a typical form of territorial power sharing, but increased


97

Binningsbø

autonomy can also be granted to specific regions only. Territorial power sharing is based on the same principles as Lijphart’s (1977) segmental autonomy, but few statistical studies take groupbased autonomy into account in their analyses. The fourth type of power sharing – economic – has not been studied as much as the other three types. Economic power sharing, often labelled wealth sharing, can be effectuated in association with territorial power sharing, where a certain region is granted both more autonomy as well as economic redistribution. This type of power sharing is often associated with natural resource-rich regions. With increasing interest in resource-related conflicts (Collier et al., 2009; Fearon and Laitin, 2003; Lujala, 2010), there has also been an increasing interest in natural resource management as a means to solve such conflicts (Binningsbø and Rustad, 2012; Le Billon and Nicholls, 2007; Lujala and Rustad, 2011). Wealth sharing is proposed as a specific type of power sharing initiated to mitigate natural resource conflicts. Both territorial and economic power sharing can be understood as arrangements aspiring to correct injustices present prior to or evolving during conflict. In the peace agreement literature, however, such provisions are introduced as means to address combatants’ security concerns. Since every researcher provides her own conceptualization of power sharing, there is an unlimited number of power sharing understandings I could discuss here. The ones elaborated on, however, represent the central conceptualizations. In particular, it is important to keep in mind the difference between long-term power sharing institutions and short-term power sharing arrangements. Long-term institutions are often formalized in constitutions, whereas transitional arrangements are mostly prescribed in peace agreements, although peace agreements may also call for long-lasting institutions.

Domain: All countries or one peace agreement? As previously outlined, Lijphart’s (1969, 1977) interest in power sharing grew out from his study of political systems in a few plural European societies. Originally, his concept of consociationalism had two dimensions, one referring to the pluralistic character of the society and one referring to elite cooperation (Lijphart, 1977). After some years Lijphart moved away from including the socio-political dimension in his concept of power sharing (1998, 2000) and in his later work discusses only the political institutions ensuring elite cooperation. Lijphart’s initial universe of cases were a few European countries – where findings could be generalized to other plural societies divided along ethnic, religious, linguistic or ideological dimensions. Which societies constitute his later universe of cases is less clear, but it does not seem to be restricted to plural societies only. Nonetheless, even if Lijphart (1998, 2000) removed pluralism as a prerequisite for consociationalism, pluralism is still considered a central aspect of the societies that will benefit most from power sharing.4 When the ‘father of power sharing’ has been ambiguous regarding which societies power sharing is best suited to, it is not too surprising that the ‘sons and daughters’ also think of power sharing as relevant for different types of societies. Overall, four categories are studied: all countries, plural societies, post-conflict societies and peace agreements. Lijphart (1977, 1985, 1996) arrived at his theory about consociational power sharing inductively, refining its theoretical and empirical characteristics as he gained new insights and discovered new cases. Unfortunately, he never made his theory or cases subject to systematic analysis.5 It is not quite clear which countries and which time periods are categorized as power sharing cases. As already pointed out, it is also not quite clear which societies that could be power sharing cases, that is, in which societies it is most relevant and which would be preferred as units of analysis.


98


Norris’s (2008) universe of cases is, on the other hand, quite clear. It consists of 191 countries during the ‘third wave’ of democratization (1974–2004). She argues – and finds – that power sharing institutions are correlated with higher levels of democracy compared with power-concentrating institutions. Her definition of power sharing does not presuppose ethnically diverse societies and her analyses look at power sharing in all societies. Much of Norris’s (2008) reasoning is, however, about power sharing as particularly helpful in plural societies and she also refers to literature on post-conflict power sharing to support her claims. Reynal-Querol (2002) also investigates all countries in the world and asks whether power sharing prevents civil war. Similarly, Schjølset (2008) studies power sharing institutions and interstate war, examining governing institutions in all countries since 1816. Neither Reynal-Querol nor Schjølset highlights the power sharing character of their studies, but they use power sharing indicators.6 Roeder (2005) compares power sharing and power dividing institutions in ethnopolitical dyads. He finds that power sharing increases the risk that the relationship between ethnopolitical dyads turns violent while power dividing reduces the risk. Roeder (2005) does not consider power sharing to be relevant for all types of countries, but focuses on the socio-political dimension of Lijphart’s (1969, 1977) original definition of consociationalism, looking only at plural societies. His units of analysis are 8074 ethnopolitical ‘dyad-quinquennia’ (5-year periods) in 153 countries between 1955 and 1999. Power sharing is increasingly used as a peace-building tool. Within this field of research the universes of cases differ between a few case studies, only signed peace agreements and all postconflict societies. While the exact definition of a post-conflict society differs among these scholars, they share the prerequisite of a violent conflict. Thus, power sharing is here understood as a governing system that is beneficial – or counter productive – for post-conflict societies only, not all countries. A central study within this field of research is Hartzell and Hoddie (2007). They investigate the effect of power sharing provisions in peace agreements; thus, their unit of analysis is not all post-conflict societies, but those emerging after negotiated settlements (see also Jarstad and Nilsson, 2008; Mattes and Savun, 2009). Walter’s (2002) and Mukherjee’s (2006b) universe of cases, on the other hand, consists of all civil wars between 1940 and 1992, but their concern is also power sharing as part of peace agreements.7 A number of researchers apply case and comparative case analyses to the study of power sharing. It can sometimes be difficult to discern the domain that these studies adhere to, but recent research seems to focus on peace agreements. The African continent is overrepresented when it comes to power sharing studies (probably also when it comes to number of power sharing peace agreements). Reynolds (1999) and Sisk and Reynolds (1998) are especially concerned with electoral rules and conflict management in African states. Additionally, Mehler (2009) critically examines power sharing provisions in peace agreements in Africa after 1999, Lemarchand (2007) studies Rwanda, Burundi and the Democratic Republic of Congo, and Tull and Mehler (2005) discuss the Democratic Republic of Congo and Cote d’Ivoire. Jarstad (2001), however, is less Africa-centred, comparing power sharing in Cyprus and New Zealand, while Sriram (2008) compares Sri Lanka, Sudan and Colombia. It is evident that power sharing researchers do not study the same types of societies. Regardless of whether one hypothesizes that power sharing works or not, research findings depend on the cases one studies. It might be true that power sharing had some perverted consequences in the Democratic Republic of Congo (Tull and Mehler, 2005); however, it also might be equally true that power sharing contributed to stability in Northern Ireland (McGarry, 2002). In addition, perhaps power sharing is not especially helpful for ethnopolitical dyads (Roeder, 2005), but it is still the best overall government for all countries (Norris, 2008).


99

Binningsbø

Causal mechanism: Ensuring minority inclusion or appeasing warlords? The causal chain linking a factor to an expected outcome is the centrepiece of every theory. This mechanism depends on both the factor and the outcome, that is, what is it exactly that is expected to bring a certain consequence. In the previous section I discussed various conceptualizations of power sharing – the factor. In this section I describe both potential outcomes of this factor and different causal mechanisms linking the two together. Jarstad (2008) argues that power sharing research can be divided into two strands of research: the ‘democracy theory’ approach focusing on power sharing as a means to achieve stable democracy – especially in divided societies – and the ‘conflict management’ approach studying power sharing’s role in achieving stable post-conflict peace. Still, although democracy theorists mainly write about stable democracy, their research is driven by the need to avoid violence in divided societies. Underlying much of the democracy theory research on power sharing, there is an assumption that heterogeneous societies are inherently conflictual. Power sharing is therefore suggested to avoid conflict and instability. While democracy theorists do not look specifically at countries damaged by civil war, it is often the case that the divisions in the countries they study have caused severe violence and perhaps even civil war in the past. Further, from a democracy theory perspective, stable democracy is viewed as a prerequisite to avoid divisions in plural societies blowing up into full-scale war in the future. Thus, lasting peace is a fundamental dependent variable also among those primarily focusing on power sharing as a means to achieve stable democracy.8 In the same way as democracy scholars are not only concerned with democracy, conflict researchers write not only about peace as a potential outcome after power sharing, but also about democratization. This is particularly true for conflict researchers critical of power sharing as a peace-building tool. Even if they acknowledge power sharing’s ability to terminate war, they argue that power sharing can be counterproductive in terms of democracy (Jarstad, 2008; Sriram and Zahar, 2009).9 As power sharing’s effect on the two potential outcomes outlined by Jarstad (2008) can be either positive or negative, there are four potential consequences of power sharing: democracy, nondemocracy, peace or armed conflict. Researchers have theorized about the causal mechanisms linking power sharing to all four outcomes, and substantiated the hypothesized relationships with empirical findings.

Power sharing causes democracy The central claim of consociational theory is that power sharing is the only alternative in order for divided societies to create stable democracy. According to Lijphart (1969, 1977), divisions among segments will be so grave that cooperation is unlikely. However, Lijphart further argues that skillful leaders will see the need for compromise and consensus to avoid society breaking down. By the aid of different power sharing institutions, in particular the grand coalition, segment leaders will cooperate and reach moderate policies all can agree to. Lijphart (1977) furthermore postulates that cooperation at the elite level over time will trickle down to cooperation at the mass level. After a while strict consociational arrangements may not be necessary anymore because the sharing of power has managed to overcome the divisions between groups in society. Consociationalism is linked to stable democracy by the claim that cooperation at the elite level brings cooperation at the mass level.10 Similar to Lijphart (1977), Norris (2008) looks at democracy as the relevant outcome for power sharing. She claims that power sharing ‘works’, that is, that power sharing mitigates and solves


100


disagreements and strengthens democracy. Norris (2008) discusses the causal links between four power sharing institutions (parliamentarism, PR elections, federalism and a free press) and quality of democracy separately. Regarding the executive, Norris argues that government’s dependence on support from parliament encourages cooperation and broad-based policy change in parliamentary systems. She further argues that PR election rule has lower hurdles for small parties, thus strengthening minority representation in decision-making bodies. Norris claims that, when groups first are represented in the legislature, they have strong incentives to negotiate and cooperate (Norris, 2008: 107). As such, PR elections will strengthen elite cooperation, but inclusion should also strengthen democratic attitudes among citizens experiencing that their group interests are represented. Additionally, Norris (2008) maintains that vertical power sharing, federalism, facilitates social stability and democratic consolidation (Norris, 2008: 157). Removing decision-making power from a national to a subnational level brings multiple access points for the citizenry. This should increase public participation in policy-making as well as making politicians more accountable and responsive (Norris, 2008: 160).11 The central causal chain between power sharing and democracy in Norris’s (2008) work is power sharing’s ability to ensure minority representation in decisionmaking processes and create incentives to cooperate.

Power sharing causes non-democracy While Lijphart (1969, 1977) and Norris (2008) claim that power sharing strengthens democracy, others are more skeptical. One seminal critique against power sharing is that it prevents a vibrant opposition. When all relevant groups are included in decision-making, opposing views are marginalized and excluded (Jung and Shapiro, 1995). The lack of strong opposition to government contradicts a central aspect of democracy where citizens can change government through election support for opposing groups. Further, pre-defined sharing of power also contradicts the principle of ex ante uncertainty (Gates and Strøm, 2013) as positions of power are not distributed according to election results, but owing to elite agreements. Following these arguments power sharing cannot bring democracy because power sharing in and of itself is undemocratic. Although Lijphart (1969, 1977) used both European and third world countries as examples of successful power sharing countries, his theory mainly builds on a few European states, in particular the Netherlands. Horowitz (1985, 1991) argues, on the other hand, that heterogeneous countries in Europe are not easily comparable with deeply divided African and Asian countries. Therefore it will be difficult to adapt Western consociational democracy in these countries, because the hostility towards members of other groups is much more intense than in the European countries (Horowitz, 1985: 572). According to Horowitz (2003), power sharing will not lead to stable democracy; in particular, he claims that PR electoral rules in divided societies strengthen ethnic divisions and magnify political differences, rather than the opposite. As mentioned above, even if power sharing research can be divided into a democracy theory approach and a conflict management approach, democracy and peace are often viewed together in much of the literature. Both Jarstad (2008) and Sriram and Zahar (2009), for example, claim that power sharing provisions in peace agreements can be counterproductive in terms of post-conflict democratization. Jarstad (2008: 106) claims that, although power sharing is prescribed as a viable solution to terminating civil war, it also has a disturbing record of bringing instability, inefficient government and even more violence to war-torn societies. According to Jarstad (2008), the key functions of power sharing can hinder long-term stability as well as instigate setbacks for future democratic rule. Power sharing is all about inclusion, but deciding who to include and exclude is not straightforward. Guaranteeing certain groups


101

Binningsbø

representation in government can make it almost impossible for other parties to get access to power later (Jarstad, 2008). Jarstad (2008: 125) further argues that power sharing at an elite level can hinder popular support for the political system since voters seem to have no influence on government. A political system without local support will also have a difficult time evolving into a wellfunctioning democracy. Similar to Jung and Shapiro’s (1995) criticism about lack of opposition, Jarstad (2008) points out how power sharing agreements reflect the power balance at the time of negotiations. Rigid rules make it harder for voters to hold politicians accountable and change government through elections, even if this would reflect popular support better. In sum, researchers critical towards power sharing as a tool to achieve democracy claim that the mechanism linking power sharing to its outcome involves power sharing strengthening group differences rather than mitigating them, preventing alternation in government, and fixing a power balance between segments in society based on the relationship at a given point in time.

Power sharing causes peace Power sharing is most often implemented in divided societies where group differences sometimes are so hostile that they develop into violent conflict. Political institutions designed to share power are expected to mitigate such conflicts. Whereas Lijphart’s (1977) original causal argument is based on elite cooperation, Jarstad (2001: 38) criticizes him for being imprecise about how this causal chain actually works. To circumvent this ambiguity, she argues that consociational theory can be better understood by two underlying theoretical dimensions: inclusion and decentralization. Jarstad (2001: 47) proposes two causal mechanisms linking the two dimensions to conflict management: inclusion levels the power relation between groups and decentralization removes conflict issues from the national political arena. Together, these two mechanisms reduce the level of conflict in society. Analogous to Jarstad’s (2001) argument about inclusion, Mukherjee (2006a) claims that democratic rule fosters enduring peace because opposing groups find themselves in environments where differences are dealt with by negotiations and non-violent bargaining. Together with formal checks and balances, and regularity of elections, this makes belligerents confident that they will not be excluded from government. Mukherjee (2006a) asserts that proportional representation electoral systems strengthen post-war peace, while majoritarian elections will not have the same impact. While Mukherjee (2006a) only indirectly discusses power sharing when he examines electoral systems, in a parallel study he (Mukherjee, 2006b) looks at political power sharing agreements and peace duration. Here he argues that insurgent leaders’ strategies are affected by their expectations of civil support, and that governments therefore will attempt to manipulate this support. Mukherjee (2006b) claims that power sharing concessions from the government weaken civil support for insurgents and consequently increase insurgent leaders’ incentive to accept the concessions (because power sharing concessions show supporters that the government considers their claims legitimate). However, Mukherjee (2006b) argues that power sharing will only be accepted by rebels if the insurgent leaders know that the government is strong and able to crush them militarily. The causal chain between power sharing and peace thus depends on a decisive victory revealing military capacity. If power sharing is initiated during military stalemate, Mukherjee (2006b) claims, it is not likely to be accepted by the rebels and thus they will not end the violence. Mukherjee’s (2006a, b) articles are two of a growing number of statistical studies on power sharing and peace that find that inclusive institutions strengthen the stability of peace after armed conflict (e.g. Hartzell and Hoddie, 2007; Jarstad and Nilsson, 2008; Mattes and Savun, 2009; Walter, 2002). While these analyses may be similar in design and conceptualization, they use somewhat different reasoning explaining the relationship.


102


Walter (2002) argues that combatants’ longer-term worries about being permanently excluded from post-conflict political power must be addressed. Contrary to the literature arguing in favour of democratic elections, she argues that peace-building initiatives that only ‘offer combatants the chance to compete in elections will not convince them to sign and implement peace agreements’ (Walter, 2002: 28, my italics). Participation has to be guaranteed. In particular, Walter (2002) considers executive power to be of utmost importance. Representation by all belligerents in government prevents former enemies from capturing the state. She further argues that in the newly established post-war government there will most likely be few constraints on the executive, and instead of being overruled by their rival in a competitive government the parties will prefer guaranteed positions in a power sharing government (Walter, 2002: 30). When such positions are offered, peace is likely to follow.12 Hartzell and Hoddie (2007; see also Hartzell, 1999; Hartzell and Hoddie, 2003; Hartzell et al., 2001; Hoddie and Hartzell, 2003) also emphasize the effect of power sharing in producing durable post-conflict peace. In particular, Hartzell and Hoddie (2007) assert the importance of power sharing along four dimensions: political, military, economic and territorial. They argue that the most extensively institutionalized settlements, that is, those with many dimensions included in peace agreements, are the ones most likely to create stable peace. Hartzell and Hoddie (2007) point to four causal chains linking extensive power sharing to peace: more dimensions are better because power sharing provides parties with influence in decision-making, and the more influence they have, the less insecure the parties will be. Further, power sharing along many dimensions will reinforce itself; when there is more than one dimension involved, groups will still be protected and have influence even if one dimension is not fully implemented; finally, extensive power sharing arrangements signal willingness to recognize and cooperate with the former enemy (Hartzell and Hoddie, 2007: 66). Mattes and Savun (2009) use – and expand – Hartzell and Hoddie’s (2007) data, but their explanation of why power sharing can bring sustainable peace is based on arguments borrowed from a bargaining model of war. They argue that civil war is costly. Thus, parties to war will try to avoid ‘the ex post inefficiency of war’ and instead ‘prefer to conclude an ex ante negotiated agreement’ (Mattes and Savun, 2009: 739). However, because the parties do not trust each other, they find it difficult to commit to an agreement that makes them vulnerable to defection by the opposing side. To overcome these commitment problems, Mattes and Savun (2009) claim that peace agreements should contain both fear-reducing and cost-increasing provisions. Provisions such as third-party guarantees, political, military, territorial and economic power sharing put constraints on the opponents’ ability to back out from the agreement, thus lessening insecurity and fear and reducing the likelihood of conflict recurrence.13 Walter (2002), Hartzell and Hoddie (2007), Mattes and Savun (2009) and Mukherjee (2006b) all find that – at least in some form – power sharing provisions in peace agreements strengthen post-conflict peace. The central causal mechanism is how power sharing overcomes commitment problems and mitigates former combatants’ (in particular rebels’) security concerns. When inclusion is guaranteed, rebels do not need to fear being marginalized in the post-conflict state. Consequently, the risk of war recurrence decreases.

Power sharing causes armed conflict While a substantial body of research argues in favour of power sharing, in terms of achieving both stable democracy and stable peace, an equally substantial body of research criticizes these claims. Causal arguments relating power sharing to non-democracy also implicitly relate power sharing to armed conflict. Similar to Horowitz’s (2003) claim that power sharing strengthens divisions rather


103

Binningsbø

than mitigating them, Rothchild and Roeder (2005: 36) argue that power-sharing institutions create ‘motives and means for the ethnic elites empowered by power sharing to escalate ethnic conflict’, and that they are ‘inflexible and unable to adapt to rapidly changing social conditions during a transition from intense conflict’ (Rothchild and Roeder, 2005: 39). Thus, power sharing does not bring peaceful relations to conflictual societies, but more violence. Mehler (2009) adds to this argument by pointing out that the elite focus of power sharing agreements ignores the crucial need for security among ordinary citizens. For Spears (2002: 127), power sharing in a post-civil war environment is to make a ‘deal with the devil’. The cooperating partner in a post-conflict society is not an ‘ordinary’ political opponent as in a stable democracy, but it is an enemy – one that has killed its opponents ‘with considerable enthusiasm and success’ (Licklider, 1995: 681). Therefore, Spears (2002: 132) argues, post-war power sharing is most likely to fail and, ironically, works best where least needed.14 According to Jarstad (2008: 110), warring parties tend to be reluctant to share power even though it is an attractive solution for negotiators. It may be easy to see why negotiators view power sharing as an attractive solution; if two – or more – groups refuse to give up a power struggle, it seems wise that both get a piece of the power pie (Jarstad, 2008). However, deciding which groups and persons to include and exclude is not straightforward. To sum up, according to the vast power sharing literature, power sharing can bring democracy, non-democracy, peace or armed conflict. The causal mechanisms applied to explain these effects highlight, for example, that inclusion ensures representation of different groups (strengthens democracy), that inclusion strengthens ethnic divisions (leads to non-democracy), that inclusion mitigates rebels’ security concerns (strengthens peace) or that inclusion rewards violent behaviour (reproduces violent conflict). Which mechanism is most ‘true’, however, depends on conceptualization of power sharing and cases studied.

Measurement: All sorts of sharing or one specific institution? The previous sections discussed power sharing in terms of conceptualization, unit of analysis and causality, showing how the understanding of power sharing – its characteristics and consequences – diverges substantially among researchers. In the following, I will discuss a more technical factor differentiating power sharing research: the indicators used to measure power sharing and the methods applied. While the relationship between concept and measurement is important in both quantitative and qualitative research, I limit this discussion to a few central quantitative power sharing studies. A logically valid link between concept and observation is necessary if one wants to draw conclusions about theoretical relationships based on empirical analyses (Adcock and Collier, 2001: 529). It is not possible to investigate the presence of ‘concepts’; instead one must operationalize concepts in order to say something meaningful about their whereabouts in society as well as their causes and consequences. How one concludes regarding how power sharing performs as a peacebuilding tool depends on which indicators one chooses. There are few, if any, published statistical studies of consociational power sharing and post-conflict peace. There is, however, a growing body of quantitative studies of power sharing peace agreements as well as power sharing and democracy. Describing the operationalizations of all these is beyond the scope of this article, but I will elaborate on three central peace agreement studies (Hartzell and Hoddie, 2007; Mukherjee, 2006b; Walter, 2002) and two power sharing studies (Mukherjee, 2006a; Norris, 2008). Details on the specific operationalizations are shown in Table 2. Walter (2002) identifies three separate dimensions of power sharing – political, military and territorial – and measures them as three dichotomous variables. Political power sharing implies


104


Table 2. Operationalizations of power sharing. Research

Power Sharing

Walter (2002: 63–64)

Political pact: combatants were offered guaranteed positions in a new government in cabinet or a higher level, or a specific quota of political power in at least one of the main branches of government Military pact: a peace settlement stipulated a quota of power in a new army Territorial pact: a peace settlement included a provision for some form of regional autonomy; one or both sides were allowed to continue to administer areas under their control; and/or specific self-governing zones were established Political power sharing: electoral, administrative, and/or executive proportional representation Territorial power sharing: federalism/confederalism and/or regional autonomy Military power sharing: the creation of state’s security forces through (a) integration of former antagonist’s armed forces on the basis of a formula representative of the size of the armed groups or equal numbers of troops; (b) appointment of members of weaker armed faction(s) to key leadership positions; (c) permission for antagonists to remain armed; and/or (d) permission for antagonists to retain their own armies Economic power sharing: resource-distribution to disadvantaged groups and/or policies to direct economic assets towards groups based on group membership or geographic location Political power sharing agreements: (a) appointment of members and/ or leaders of insurgent groups to ministerial, sub-ministerial or cabinet positions; (b) appointment of members and/or leaders of insurgent groups to central bureaucracy, civil service, courts, foreign service and commissions; (c) use of proportional electoral systems and/or (d) allowance of members of insurgent groups to organize and form political parties and participate in elections Political regime: democracy, autocracy (Boix, 2003; Marshall and Jaggers, 2003; Przeworski et al., 2000) Type of executive: parliamentary democracy, presidential democracy (Delury, 1999; Derbyshire and Derbyshire, 2000; Kurian, 1998; Przeworski et al., 2000) Electoral system: PR elections, majoritarian elections (Delury, 1999; Derbyshire and Derbyshire, 2000; Kurian, 1998) Electoral system: proportional representation, majority-plurality, and combined electoral rules (Reynolds et al., 2005). Positive action strategies: for example, electoral boundaries specifically designed to ensure minority representation or reserved seats (Norris, 2008: 118) Types of executive: ruling monarchies, parliamentary monarchies, presidential republics, mixed republics and military states (Banks, 2000) Federalism: federal constitution, decentralized unions (hybrid states), unitary states (various handbooks, e.g. Elazar, 1994; Watts, 1999) Press freedom: free, partly free, not free (Freedom House, various years)

Hartzell and Hoddie (2007: 159)

Mukherjee (2006b: 494)

Mukherjee (2006a) Norris (2008)


105

Binningsbø

including former rebels in a joint government; military power sharing can allow combatants to retain their defenses, for example by including equal numbers of rebel and government soldiers in a new national army. The last type of power sharing concerns territorial control, for example by ensuring that parties to a civil war gain administrative control over occupied territories (Walter, 2002). The dependent variable in Walter’s (2002) study is a categorical variable with four outcomes (no negotiation, active formal negotiation, signed bargain and successfully implemented settlement), which is analysed using an ordered logit model. By examining 72 civil wars initiated between 1940 and 1992, Walter (2002: 84) finds that a peace agreement is 18% more likely to be successfully implemented if there is a territorial power sharing pact and 16% more likely if there is a political power sharing pact.15 Hartzell and Hoddie (2007) add a fourth, economic, dimension to the three power sharing dimensions included in Walter’s (2002) study. Also different from Walter (2002), Hartzell and Hoddie’s (2007) independent variable is ‘settlement institutionalization’ and is a numeric variable taking numbers from 0 to 4, depending on how many dimensions of power sharing were included in a negotiated settlement. Hartzell and Hoddie (2007) use a Cox proportional hazards model with the duration of peace in months since conflict termination as a dependent variable. Analysing 38 negotiated settlements and 11 negotiated truces of civil wars that broke out between 1945 and 1999, they find strong support for their expectation that extensively institutionalized settlements produce long-lasting peace (Hartzell and Hoddie, 2007: 75).16 Mukherjee (2006b) takes Walter (2002) and Hartzell and Hoddie (2003) as a starting point, but limits himself to political power sharing only. He argues that one should not look at power sharing in isolation, but rather in interaction with type of conflict termination. Similar to Hartzell and Hoddie (2007), Mukherjee (2006b) uses a Cox proportional hazards model with peace duration in months as a dependent variable. He analyses 111 civil wars between 1944 and 1999 and finds that the interaction between victory and political power sharing prolongs peace duration, while the interaction between military stalemate and political power sharing shortens peace duration. All his independent variables are dichotomous.17 In addition to his study of political power sharing agreements, Mukherjee (2006a) also investigates the effect of political institutions on post-conflict peace duration. He does not use the words ‘power sharing’ to describe the institutions he studies; however, he bases his hypotheses on Lijphart’s (1977) consociational theory. Thus, both studies can be viewed as tests of the effect of post-conflict power sharing on peace. In Mukherjee (2006a) he analyses two power sharing political institutions: parliamentarism and PR elections. In this study Mukherjee (2006a) examines the same 111 civil wars as described above, but uses a Weibull duration model rather than Cox. The dependent variable is peace duration in months, while the independent ones are dummy variables. He finds that parliamentary democracy and PR electoral rules significantly decrease the risk of post-conflict peace failure.18 A peace-strengthening effect of power sharing is thus confirmed also with Mukherjee’s (2006a) analyses. Norris (2008) investigates some of the same institutions as Mukherjee (22006a); however, she clearly considers these institutions to be power sharing institutions, which is also illustrated in the sub-title of her book: Does Power-Sharing Work? In addition to parliamentarism and PR elections, Norris (2008) looks at federalism and a free press. She compares these power sharing institutions, measured as dichotomous variables, with other types of (power-concentrating) governing institutions. Norris’s (2008) dependent variables are various measures of democracy and she uses ordinary least squares regressions to investigate the effect of power sharing on these.19 Her analyses find strong support for the hypothesis that power sharing institutions work, that is, that such institutions are associated with higher democracy scores than other types of institutions.


106


Table 3. Power sharing in four post-conflict societiesa.

Walter (2002) Hartzell and Hoddie (2007) Mukherjee (2006b) Mukherjee (2006a) Norris (2008)c

Political pact Territorial pact Military pact Political power sharing Territorial power sharing Military power sharing Economic power sharing Political power sharing Democracy Parliamentarism PR Parliamentary monarchy PR Positive action strategies Federal constitution Free pressd

Cambodia (1978–1998)

Guatemala (1965–1995)

Mozambique (1977–1992)

Sierra Leone (1991–2000)

1 0 0 1

0 0 1 0

1 0 1 1

n.a. n.a. n.a. 1

0

1

1

0

1

0

1

1

0

1

0

1

1

0

1

n.a.

0 0 1b 1

1 0 1 0

0 0 1 0

1 0 0 0

1 0

1 0

1 0

1 0

0

0

0

0

2

6

1

5

aConflict years are from UCDP/PRIO Armed Conflict Data (Gleditsch et al., 2002); however, because of different coding criteria for conflict, conflict years vary across the other datasets. In Hartzell and Hoddie’s (2007), Mukherjee’s (2006a, b) and Walter’s (2002) data, the Cambodian civil war ends in 1991. The Guatemalan civil war ends in 1996 according to Hartzell and Hoddie (2007) and Walter (2002), while the Sierra Leonean civil war ends in 1999 according to Hartzell and Hoddie (2007) and Mukherjee (2006a, b). bIn Mukherjee (2006a) the score for PR changes in Cambodia; it is 0 until 1993 and 1 thereafter. cScores on Norris’s (2008) indicators are from the first post-conflict year, except for Mozambique, with scores for 1994. dFreedom House categories of free press: 0–30 ‘free’; 31–60 ‘partly free’; and 61–100 ‘not free’.n.a., Not applicable.

A simplified, overall conclusion about power sharing and peace from these studies is that power sharing increases the chances of post-conflict peace and strengthens democracy.20 A comparison of the three studies of power sharing provisions in peace agreements shows quite diverging approaches to the question. Walter (2002) applies the narrowest operationalization of political power sharing, requiring a political pact where combatants are offered guaranteed positions or specific quotas in cabinet (or a different branch of government). Hartzell and Hoddie (2007), however, in addition to political power sharing as operationalized in Walter (2002), also include electoral proportionality; thus it is not necessary that the power sharing involves the combatants. Mukherjee (2006b) broadens the operationalization even more, also defining peace settlements where former rebels are allowed to transform into a political party and participate in elections as power sharing. Thus, he does not require that the combatants win a single seat nor participate in government, only that they


107

Binningsbø

are allowed to. The operationalizations of military power sharing in Walter (2002) and Hartzell and Hoddie (2007) are more similar, although Hartzell and Hoddie are more specific in their listings. Territorial power sharing is the same in both Walter (2002) and Hartzell and Hoddie’s (2007) studies. Hartzell and Hoddie (2007) are the only ones to include economic power sharing. Regarding the political institutions analysed in Mukherjee (2006a) and Norris (2008), it is more difficult to disclose differences in coding criteria, as they, in theory, measure the same type of institutions. However, Mukherjee’s (2006a) data are less fine-tuned than Norris’s (2008) data, as he separates between fewer categories. This can explain why there are some differences in some countries’ scores for the power sharing institution variables. Ideally, we should know the correlations between these different power sharing measurements, but because of the substantial variance in domain (peace agreements, civil wars and all countries), it is difficult to gather a large enough set of comparable cases to calculate a correlation coefficient. Table 3 presents the available scores for the different power sharing indicators for four post-conflict countries: Cambodia, Guatemala, Mozambique and Sierra Leone. From this small sample it is clear that power sharing is not measured similarly across different datasets. For example, Hartzell and Hoddie (2007) code territorial power sharing in postconflict Guatemala and Mozambique, while Walter (2002) does not. Walter (2002), however, codes military power sharing in Guatemala, something Hartzell and Hoddie (2007) do not. Norris (2008) codes PR election in Sierra Leone, but Mukherjee (2006a) does not. Additionally, Norris (2008) and Mukherjee (2006a) disagree over the coding of parliamentarism in Cambodia – Norris (2008) claims the country is a parliamentary system, while Mukherjee (2006a) claims it is not. These differences most likely influence results from statistical analyses. When analysing smaller samples – such as analyses of only peace agreements (Hartzell and Hoddie, 2007) – the scores on one variable for one specific case can influence the effect of that variable substantially. Researchers, and policy-makers listening to researchers’ advice, must take this into account. In addition to differences in conceptualization, domain, causality and measurement, analyses of power sharing vary in terms of definition of dependent variable, inclusion of control variables and methods applied. Different decisions regarding these aspects influence analyses and findings about power sharing too, even if the core definitions are the same.

Concluding remarks Power sharing is a governing system aiming at including multiple political actors in decisionmaking processes (Norris, 2008). The previous sections have shown that such inclusion can be arrived at through different means. Power sharing also aims at reducing the level of conflict in a society, be it violent or not. Thus, power sharing is understood and analysed as a governing system aiming at establishing democracy and/or sustainable peace. However, whether one can conclude that power sharing actually achieves these goals – or not – depends on the conceptualization of power sharing, where it is considered relevant, which causal mechanisms are at play and the analytical approach to the topic. This article shows the diversity when it comes to understandings of power sharing. It shows that the conceptualization of power sharing varies from ‘government by elite cartel’ (Lijphart, 1969: 216) institutionalized in a grand coalition, mutual veto, proportionality and segmental autonomy (Lijphart, 1977), via parliamentary systems, PR elections, federalism and a free press (Norris, 2008), to, for example, rebel inclusion in a new national army (Hartzell and Hoddie, 2007; Hoddie


108


and Hartzell, 2003). In addition, these different conceptualizations are considered relevant for different types of societies. In some instances, power sharing is investigated as a governing system potentially relevant for all countries in the world (Norris, 2008), while in other studies it is only analysed in divided societies (Roeder, 2005). As a peace-building tool, power sharing is investigated in all post-conflict societies (Mukherjee, 2006a, b; Walter, 2002), or only as provisions in peace agreements (Hartzell and Hoddie, 2007; Jarstad and Nilsson, 2008; Mehler, 2009; Tull and Mehler, 2005). More important, since power sharing researchers disagree on the consequences of power sharing – it can bring democracy, non-democracy, peace or violence – the causal chains relating power sharing to its outcome differ as well. Researchers claiming that power sharing brings democracy argue that it ensures minority representation and closeness to decision-making (Norris, 2008), while the counter-argument is that power sharing strengthens divisions and exacerbate disagreements (Rothchild and Roeder, 2005). Power sharing is further theorized to bring peace because it addresses post-conflict security concerns by ensuring that rebel groups are guaranteed positions of power (Hartzell and Hoddie, 2007; Walter, 2002); however, it may also increase the risk of conflict because it rewards violent behaviour (Tull and Mehler, 2009). Finally, research on power sharing applies different operationalizations of the concept. The measurements of power sharing can be limited to provisions in peace agreements that ensure rebel inclusion in legislative or executive branches of government (Jarstad, 2008), or can be wide enough to include rebel groups’ transformation to political parties (Mukherjee, 2006b). There may not be one ‘true’ power sharing or one ‘truth’ about power sharing. However, it is crucial that researchers and policy-makers are clear about which type of power sharing they discuss in specific situations. Given that power sharing is increasingly recommended and implemented in many fragile and post-conflict societies, it is important to understand what is meant by power sharing in these contexts. Acknowledgments I am grateful to Kristin M. Bakke, Michael Bratton, Scott Gates, Anna Jarstad, Timothy Sisk, Kaare Strøm and the anonymous reviewers for helpful comments, and to Andreas Mehler for inspiration regarding the title.

Funding The research was supported by the Research Council of Norway (grant no 196850) and the Norwegian Ministry of Foreign Affairs.

Notes 1. Adding to the confusion, Svolik (2012) refers to authoritarian coalitions as power sharing. Such arrangements are not covered in this article. 2. Lijphart later introduced the concept of ‘power-sharing democracy’ as a synonym, power sharing being not so ‘much of a tongue-twister’ (Lijphart, 1985, 1998: 100). Lemarchand (2007: 3), however, argues that power sharing can be viewed as ‘ad hoc concessions’ aiming at co-opting rebels, while consociationalism implies carefully designed constitutional norms. 3. However, some of the operationalizations of political power sharing only enable rebel representation, and do not guarantee it. 4. Even in studies of power sharing looking at all countries in the world, the reasoning presented relies on power sharing as especially suitable for ethnically, religiously and/or linguistically divided societies. 5. In 1999 Lijphart introduced ‘consensus democracy’ as a political system in contrast to majoritarian governance. He analysed this type of government looking at 36 stable democracies and applying factor analysis to find similarities and dissimilarities (Lijphart, 1999).


109

Binningsbø

6. Both use the concept of power sharing in their work and refer to the most central power sharing literature (such as Lijphart, 1977, 1999), but neither Reynal-Querol (2002) nor Schjølset (2008) seem to consider their own research as part of the power sharing literature. 7. Both Walter (2002) and Hartzell and Hoddie (2007) add case studies to their otherwise quantitative studies. Walter (2002) compares Zimbabwe (success) with Rwanda (failure), while Hartzell and Hoddie (2007) compare the Philippines (success) with Angola (failure). 8. Norris (2008), for example, refers to post-conflict research on power sharing to support her claims about power sharing and democracy. 9. Quantitative conflict researchers, on the other hand, are more likely to focus on power sharing’s effect on peace only (Hartzell and Hoddie, 2007; Jarstad and Nilsson, 2008; Mattes and Savun, 2009; Mukherjee, 2006a, b, Walter, 2002). 10. Andeweg (2000: 520), however, points out that claiming that ‘elite cooperation leads to political stability’ is close to a tautology and criticizes Lijphart for not really elaborating on the causal chain linking elite cooperation to societal cooperation. 11. Norris (2008: 186) also claims that an independent mass media is associated with higher levels of democracy. 12. Walter (2002) argues that, in addition to power sharing, combatants need third parties to guarantee their security while they disarm and demobilize. These outside actors must provide forces to reassure the weaker party ‘that it will not be crushed and deter the stronger side from exploiting its advantages’ (Walter, 2002: 27). 13. Additionally, arrangements that increase the cost of fighting, such as provisions about withdrawal of foreign forces, border seals, separation of troops and peacekeeping operations, make it less likely that former combatants will return to military campaigns (Mattes and Savun, 2009: 738). 14. Kaufmann (1996) has also criticized the power sharing system for not being able to handle grave contradictions. It may perhaps prevent potential ethnic conflict and dampen mild conflicts, but falls short of bringing peace in situations with intense violence. 15. The most central factor in Walter’s analysis of post-conflict peace is, however, outside security guarantees. She finds that ‘even the most detailed power-sharing arrangements were not enough to ensure a successfully negotiated settlement in the absence of third-party guarantees’ (Walter, 2002: 86). 16. Mattes and Savun (2009) use an updated version of Hartzell and Hoddie’s (2007) data, but analyse the four dimensions separately. They find no significant effect of territorial, military or economic power sharing, only that political power sharing reduces the risk of post-conflict peace failure. 17. Mukherjee’s (2006b) analyses confirm his claims as none of his independent variables (political power sharing agreement, government victory, insurgent victory and military stalemate) have significant individual effects on peace duration. 18. Presidential democracy also reduces the risk of peace failure; however, the effect is not as strong as for parliamentary systems. Majoritarian electoral rules have no significant effect. 19. Norris (2008) tests the effect of governing institutions on democratic performance using four different measures of democracy (Freedom House, various years; Marshall and Jaggers, 2003; Przeworski et al., 2000; Vanhanen, 2000). 20. Jarstad and Nilsson (2008) examine both provisions of power sharing in peace agreements and the implementation of these provisions. They find that territorial power sharing provisions reduce the risk of peace failure, as similarly do implementation of territorial and military power sharing. Neither political pacts nor the implementation of such have any significant effect. Their dataset is unfortunately not available for replication.

References Adcock R and Collier D (2001) Measurement validity: A shared standard for qualitative and quantitative research. American Political Science Review 95(3): 529–546. Andeweg RB (2000) Consociational democracy. Annual Review of Political Science 3: 509–536.


110


Banks AS (2000) Cross-polity Time-series Database. Binghamton, NY: State University of New YorkBinghamton. Binningsbø HM and Rustad SA (2012) Sharing the wealth: A pathway to peace or a trail to nowhere? Conflict Management and Peace Science 29(5): 497–516. Bogaards M (2000) The uneasy relationship between empirical and normative types in consociational theory. Journal of Theoretical Politics 12(4): 395–423. Boix C (2003) Democracy and Redistribution. Cambridge: Cambridge University Press. Collier P, Hoeffler A and Rohner D (2009) Beyond greed and grievance: Feasibility and civil war. Oxford Economic Papers 61(1): 1–27. Delury GE (1999) World Encyclopedia of Political Systems and Parties. New York: Facts on File. Derbyshire D and Derbyshire I (2000) Encyclopedia of World Political Systems. Armonk, NY: Sharpe. Elazar D (1994) Federal Systems of the World: A Handbook of Federal, Confederal and Autonomy Arrangements. Essex: Longman. Fearon JD and Laitin D (2003) Ethnicity, insurgency, and civil war. American Political Science Review 97(1): 75–90. Freedom House (various years) Freedom in the world, www.freedomhouse.org (accessed 28 November 2012). Gates S and Strøm K (2013) The appeals and perils of power-sharing. In: Gates S and Strøm K (eds) Fragile Bargains: Civil Conflict and Power-sharing in Africa (in press). Gilbert ER (2009) Ethnic or non-ethnic power sharing? Addressing the issue of ethnicity in African power sharing agreements. In: AEGIS Thematic Conference on Power Sharing Agreements in Africa: Implications for Peace, Democracy and Societal Trust, Hamburg, 1–2 October. Glassmyer K and Sambanis N (2008) Rebel-military integration and civil war termination. Journal of Peace Research 45(3): 365–384. Gleditsch NP, Wallensteen P, Eriksson M, Sollenberg M and Strand H (2002) Armed conflict 1946–2001: A new dataset. Journal of Peace Research 39(5): 615–637. Hartzell CA (1999) Explaining the stability of negotiated settlements to intrastate wars. Journal of Conflict Resolution 43(1): 3–22. Hartzell CA and Hoddie M (2003) Institutionalizing peace: Power sharing and post-civil war conflict management. American Journal of Political Science 47(2): 318–332. Hartzell CA and Hoddie M (2007) Crafting Peace. Power-sharing Institutions and the Negotiated Settlement of Civil Wars. University Park, PA: Pennsylvania State University Press. Hartzell CA, Hoddie M and Rothchild D (2001) Stabilizing the peace after civil war: An investigation of some key variables. International Organization 55(1): 183–208. Hoddie M and Hartzell CA (2003) Civil war settlements and the implementation of military power-sharing arrangements. Journal of Peace Research 40(3): 303–320. Horowitz DL (1985) Ethnic Groups in Conflict. Berkeley, CA: University of California Press. Horowitz DL (1991) A Democratic South Africa? Constitutional Engineering in a Divided Society. Berkeley, CA: University of California Press. Horowitz DL (2003) The cracked foundations of the right to secede. Journal of Democracy 14(2): 5–17. Jarstad AK (2001) Changing the game: Consociational theory and ethnic quotas in Cyprus and New Zealand. Doctoral thesis, Department of Peace and Conflict Research, Uppsala University. Jarstad AK (2008) Power sharing: Former enemies in joint government. In: Jarstad AK and Sisk TD (eds) From War to Democracy: Dilemmas of Peacebuilding. Cambridge: Cambridge University Press, pp. 105–133. Jarstad AK and Nilsson D (2008) From words to deeds: The implementation of power-sharing pacts in peace accords. Conflict Management and Peace Science 25(3): 206–223. Jung C and Shapiro I (1995) South Africa’s negotiated transition: Democracy, opposition, and the new constitutional order. Politics and Society 23(3): 269–308. Kaufmann C (1996) Possible and impossible solutions to ethnic civil wars. International Security 20(4): 136–175.


111

Binningsbø

Kurian GT (1998) World Encyclopedia of Parliaments and Legislatures. Washington, DC: Congressional Quarterly. Le Billon P and Nicholls E (2007) Ending ‘resource wars’: Revenue sharing, economic sanction or military intervention? International Peacekeeping 14(5): 613–632. Lemarchand R (2007) Consociationalism and power sharing in Africa: Rwanda, Burundi and the Democratic Republic of Congo. African Affairs 106(422): 1–20. Licklider R (1995) The consequences of negotiated settlements in civil wars, 1945–1993. American Political Science Review 89(3): 681–690. Lijphart A (1968) Typologies of democratic systems. Comparative Political Studies 1(1): 3–44. Lijphart A (1969) Consociational democracy. World Politics 21(2): 207–225. Lijphart A (1977) Democracy in Plural Societies. New Haven, CT: Yale University Press. Lijphart A (1985) Power-sharing in South Africa. Institute of International Studies, University of California, Berkeley. Lijphart A (1996) The puzzle of Indian democracy: A consociational interpretation. American Political Science Review 90(2): 258–268. Lijphart A (1998) Consensus and consensus democracy: Cultural, structural, functional, and rational-choice explanations. Scandinavian Political Studies 21(2): 99–108. Lijphart A (1999) Patterns of Democracy: Government Forms and Performance in Thirty-six Countries. New Haven, CT: Yale University Press. Lijphart A (2000) Definitions, evidence, and policy: A response to Matthijs Bogaards’ Critique. Journal of Theoretical Politics 12(4): 425–431. Lijphart A (2002) Negotiation democracy versus consensus democracy: Parallel conclusions and recommendations. European Journal of Political Research 41(1): 107–113. Lujala P (2010) The spoils of nature: Armed civil conflict and rebel access to natural resources. Journal of Peace Research 47(1): 15–28. Lujala P and Rustad SA (eds) 2011. High-value Natural Resources and Post-conflict Peacebuilding. London: Earthscan. Marshall MG and Jaggers K (2003) Polity IV Project: Political regime characteristics and transitions, 1800–2002: Dataset users’ manual. University of Maryland, http://www.nd.edu/~mcoppedg/crd/ PolityIVUsersManualv2002.pdf (accessed 28 November 2012). Mattes M and Savun B (2009) Fostering peace after civil war: Commitment problems and agreement design. International Studies Quarterly 53(3): 737–759. McGarry J (2002) ‘Democracy’ in Northern Ireland: Experiments in self-rule from the protestant ascendancy to the Good Friday agreement. Nations and Nationalism 8(4): 451–474. Mehler A (2009) Peace and power sharing in Africa: A not so obvious relationship. African Affairs 108(432): 453–473. Mukherjee B (2006a) Does third-party enforcement or domestic institutions promote enduring peace after civil wars? Policy lessons from an empirical test. Foreign Policy Analysis 2(4): 405–430. Mukherjee B (2006b) Why political power-sharing agreements lead to enduring peaceful resolution of some civil wars, but not others? International Studies Quarterly 50(2): 479–504. Norris P (2008) Driving Democracy: Do Power-sharing Institutions Work? Cambridge: Cambridge University Press. Przeworski A, Alvarez ME, Cheibub A and Limongi F (2000) Democracy and Development: Political Institutions and Well-being in the World, 1950–1990. Cambridge: Cambridge University Press. Reilly B (2005) Does the choice of electoral system promote democracy? The gap between theory and practice. In: Roeder PG and Rothchild D (eds) Sustainable Peace: Power and Democracy after Civil Wars. Ithaca, NY: Cornell University Press, pp. 159–171. Reynal-Querol M (2002) Ethnicity, political systems, and civil wars. Journal of Conflict Resolution 46(1): 29–54. Reynolds A (1999) Electoral Systems and Democratization in Southern Africa. Oxford: Oxford University Press.


112


Reynolds A, Reilly B and Ellis A (2005) Electoral System Design: The New International IDEA Handbook. Stockholm: International Institute for Democracy and Electoral Assistance. Roeder PG (2005) Power dividing as an alternative to ethnic power sharing. In: Roeder PG and Rothchild D (eds) Sustainable Peace: Power and Democracy after Civil Wars. Ithaca, NY: Cornell University Press, pp. 51–82. Roeder PG and Rothchild D (eds) (2005) Sustainable Peace: Power and Democracy after Civil Wars. Ithaca, NY: Cornell University Press. Rothchild D and Roeder PG (2005) Power sharing as an impediment to peace and democracy. In: Roeder PG and Rothchild D (eds) Sustainable Peace. Power and Democracy after Civil Wars. Ithaca, NY: Cornell University Press, pp. 29–50. Schjølset A (2008) Institutional variance of the democratic peace, 1816–2002: Electoral, executive, and federal systems in time and space. Doctoral thesis, International Relations, University of Southern California. Schneckener U (2002) Making power-sharing work: Lessons from successes and failures in ethnic conflict regulation. Journal of Peace Research 39(2): 203–228. Sisk TD (1995) Democratization in South Africa: The Elusive Social Contract. Princeton, NJ: Princeton University Press. Sisk TD and Reynolds A (eds) (1998) Elections and Conflict Management in Africa. Washington, DC: United States Institute of Peace. Spears IS (2000) Understanding inclusive peace agreements in Africa: The problems of sharing power. Third World Quarterly 21(1): 105–118. Spears IS (2002) Africa: The limits of power-sharing. Journal of Democracy 13(3): 123–136. Sriram CL (2008) Peace as Governance. Power-Sharing, Armed Groups and Contemporary Peace Negotiations. London: Palgrave Macmillan. Sriram CL and Zahar MJ (2009) The perils of power-sharing: Africa and beyond. Africa Spectrum 44(3): 11–39. Svolik MW (2012) The Politics of Authoritarian Rule. Cambridge: Cambridge University Press. Tull DM and Mehler A (2005) The hidden costs of power-sharing: Reproducing insurgent violence in Africa. African Affairs 104(416): 375–398. Vanhanen T (2000) A new dataset for measuring democracy, 1810–1998. Journal of Peace Research 37(2): 251–165. Walter BF (2002) Committing to Peace: The Successful Settlement of Civil Wars. Princeton, NJ: Princeton University Press. Watts RL (1999) Comparing Federal Systems. Kingston: McGill-Queen’s University Press.


Power sharing, peace and democracy: Any obvious relationships?

Democracy and Power-Sharing in Stormy Weather

Driving Democracy: Do Power-Sharing Institutions Work?

Peace as Governance: Power-Sharing, Armed Groups and Contemporary Peace Negotiations (Rethinking Peace and Conflict Studies)

Sharing Executive Power: Roles and Relationships at the Top

Peace as Governance: Power-Sharing, Armed Groups and Contemporary Peace Negotiations (Rethinking Peace and Conflict Studies)

Peace as Governance: Power-Sharing, Armed Groups and Contemporary Peace Negotiations (Rethinking Peace and Conflict Studies)

Gender, Power and Relationships

Gender, Power and Relationships

Sistering: Power and Change in Female Relationships

Obvious Factor

Thinking about Democracy: Power Sharing and Majority Rule in Theory and Practice

Crafting Peace: Power-Sharing Institutions and the Negotiated Settlement of Civil Wars

Intelligence Power in Peace and War

Intelligence Power in Peace and War

The Power of Peace (Journal)

Syndromes of Corruption: Wealth, Power, and Democracy

Everything Is Obvious

Everything Is Obvious

Power-Sharing: Institutional and Social Reform in Divided Societies

Must Global Politics Constrain Democracy? Great-Power Realism, Democratic Peace, and Democratic Internationalism

Must Global Politics Constrain Democracy? Great-Power Realism, Democratic Peace, and Democratic Internationalism

Must Global Politics Constrain Democracy? Great-Power Realism, Democratic Peace, and Democratic Internationalism

Democracy and Ethnic Conflict: Advancing Peace in Deeply Divided Societies

Relationships

The Democracy Advantage: How Democracies Promote Prosperity and Peace

Sharing Shannon

Sharing Shannon

Sharing Sunrise

Power sharing, peace and democracy: Any obvious relationships?

Democracy and Power-Sharing in Stormy Weather

Driving Democracy: Do Power-Sharing Institutions Work?

Peace as Governance: Power-Sharing, Armed Groups and Contemporary Peace Negotiations (Rethinking Peace and Conflict Studies)

Sharing Executive Power: Roles and Relationships at the Top

Peace as Governance: Power-Sharing, Armed Groups and Contemporary Peace Negotiations (Rethinking Peace and Conflict Studies)

Peace as Governance: Power-Sharing, Armed Groups and Contemporary Peace Negotiations (Rethinking Peace and Conflict Studies)

Gender, Power and Relationships

Gender, Power and Relationships

Sistering: Power and Change in Female Relationships

Obvious Factor

Thinking about Democracy: Power Sharing and Majority Rule in Theory and Practice

Crafting Peace: Power-Sharing Institutions and the Negotiated Settlement of Civil Wars

Intelligence Power in Peace and War

Intelligence Power in Peace and War

The Power of Peace (Journal)

Syndromes of Corruption: Wealth, Power, and Democracy

Everything Is Obvious

Everything Is Obvious

Power-Sharing: Institutional and Social Reform in Divided Societies

Must Global Politics Constrain Democracy? Great-Power Realism, Democratic Peace, and Democratic Internationalism

Must Global Politics Constrain Democracy? Great-Power Realism, Democratic Peace, and Democratic Internationalism

Must Global Politics Constrain Democracy? Great-Power Realism, Democratic Peace, and Democratic Internationalism

Democracy and Ethnic Conflict: Advancing Peace in Deeply Divided Societies

Relationships

The Democracy Advantage: How Democracies Promote Prosperity and Peace

Sharing Shannon

Sharing Shannon

Sharing Sunrise

Recommend Documents